Jun Qi, Zhuoqian Zhao, Zhen Luo, Junwei Sun, J. Y. Yang, Haixin Wang, Yubo Liu
The integration of electric vehicles into decentralized energy sharing systems demands innovative parallel computing solutions to address the real-time processing and trust challenges in vehicle-to-vehicle transactions. This study tackles two critical limitations: (1) conventional routing methods’ inability to handle dynamic spatio-temporal constraints, and (2) centralized reputation mechanisms conflicting with vehicle-to-vehicle’s decentralized nature. This paper proposes a parallel distributed computing framework combining spatio-temporal network optimization with blockchain architecture. First, a tensor-based spatio-temporal network model converts dynamic routing into parallelizable static flow allocation, enabling real-time constraint embedding through distributed parallel matrix operations. Second, a blockchain-powered transaction layer implements parallel smart contracts for concurrent verification and social welfare-optimized pricing, achieving Byzantine fault-tolerant consensus through sharded transaction processing. Experimental results demonstrate 23.6% faster computation throughput and 31.2% higher transaction concurrency compared to existing distributed systems, while maintaining 17.8% social welfare improvement. The framework effectively resolves the latency-trust dichotomy in V2V energy exchange through coordinated parallel computing paradigms.